Epigenome-wide association of response to treatment with electroconvulsive therapy (ect)

Major unipolar or bipolar depression disorders are common and serious medical conditions associated with reduced quality of life and increased mortality. About 30% of patients do not respond to standard treatments, which include psychosocial, cognitive, and pharmacotherapeutic approaches. However, 70-90% of these treatment-resistant patients have been shown to respond to electroconvulsive therapy (ECT). During ECT, electric currents are passed through the brain to trigger a brief seizure. Successful treatment with ECT leads to improved neurocognitive functioning, although side effects, such as loss of autobiographic memories, are common. ECT has remained the most effective acute treatment for major depressive episodes over the past 80 years, yet the mechanisms for its effects are poorly understood. While experienced clinicians can predict who will respond to the treatment, it seems more difficult to predict who will not respond. A biological marker that can differentiate between those who will respond and those who will not is needed. Epigenetic mechanisms respond to environmental stimuli and may therefore be a suitable readout for gaining knowledge about the mechanisms involved in ECT treatment, and treatment response. We have therefore typed DNA methylation (DNAm) profiles and carried out an epigenome-wide association study (EWAS) in blood from a Norwegian cohort (N=65). Blood was drawn from patients before and after ECT treatment, and the samples were typed using Illumina EPIC arrays and preprocessed using standard protocols. We applied a mixed linear regression approach, as implemented in the Limma R package, to identify cytosines (CpGs) associated with 1) response (change in depression score), 2) treatment (change in DNAm before and after ECT) and 3) the interaction between response and treatment. We also performed a meta-analysis using the results from the EWAS on this cohort and a previously published study on a German cohort (N=34) (Sirignano et al. Trans. Psych 2021). Comb-p was used to find differentially methylated regions (DMRs) and the R package Meta was employed for the meta-analysis. While the treatment effect is modest in this study, we did identify several significant DMRs associated with treatment response, one of which is overlapping in the Norwegian and German cohorts.